How large-scale AI brands are helping to empathize with consumers.
Last week, cross-channel marketing platform Iterable announced the launch of Brand Affinity, a new cross-channel personalization tool based on customer sentiment analysis as its primary analysis. Simply put, by using AI to measure customer sentiment, the solution can create brand segments that can be scaled appropriately.
By 2020, marketers should see customer sentiment like never before: lives have changed, priorities have changed, buying patterns have changed, and consumers are buying with purchasing power in a digital environment. Infinity, not just on the highway.
It was a great time to see how innovative brands are helping us understand how we feel. First, we looked for Timetable’s VP, Product Bella Step nova, to learn more about its relationship with the brand.
A positive state of mind
Because it is a platform designed to deliver connected customer experiences across all channels, repeatable brands such as Ipsy, Door Dash, and Curology can communicate with their customers in a variety of digital configurations, rather than through a static eCommerce website. “Most of our customers try to start a good conversation between the channels so that they can easily switch between the channels,” confirms Stepanova.
Personal solutions are of course widely offered, but the relationship with the brand emphasizes empathy. ‘When I think about communication and customer experience, empathy is everything. When a business is small, it is very easy to understand what the customer is feeling because he knows all the customers. Once it reaches millions of customers and receives billions of signals, it’s not something we can handle without AI.
Iterable’s AI analyzes this vast amount of data to analyze sentiment and how it changes over time for each individual. Sentiment tags are attached to the customer’s records in a negative-to-true spectrum. Marketers can use this continuously updated field in the registry to reach their target audience. Messages are tailored to sentiment-defined segments, not individual customers.
Brand Affinity is committed to going beyond the tools to predict abandonment and identify positive and negative mood customers. “There are more use cases,” said Stepanova. “Clients with negative feelings often lose hope, but in addition to the risk of rewards, you can also look at sentiment programs.”
Satisfied customers may, for example, be open to incremental or cross-selling. For example, a customer who repeats and uses beta brand affinity can identify users who want to move from loyal donors to nonprofit sites. “By addressing the right audience and understanding loyal people who want to become donors, they can double the conversion.”
Founded in 2013 by Justin Zhu and Andrew Boni, former Twitter and Google engineers, Iterable won $ 60 million in Series D in December 2019.
Look at the whole picture
Matt Nolan, senior director of product marketing for marketing, artificial intelligence, and decision making at Pega, has a long-term view of where marketing comes from and where. When two people talk to each other, especially face to face, they use all kinds of unconsciously received signals to understand how the other person is feeling. “Most of the software is like a human being without intuition: you can’t see anyone, so we have to train them.”
Marketers have been working on this challenge since the inception of direct marketing and mailing list filtering, using increasingly sophisticated business rules-based tools. This was the beginning of the segmentation. According to him, 15 years ago it was the last product, the second-best product model, that identified the customer groups that would likely be interested in a product, depending on the purchasing model.
Ten years ago: “Let’s take a look at all the different conversations we want to have with a customer. Not just the products we want to sell, but also other conversations that can add more value, such as There is an act of service that you want the “selling” to be proactive. “Are there ways to encourage you to build a better relationship?
This was the second-best approach. “Where the market is now,” Nolan continues, “is empathy. Try to truly understand the customer, put yourself in their shoes. What has changed is the availability and speed of data. There is a constant flow. Brand information from customers. Provides their signals that can be interpreted to find out if there is a way to help, support or market it. It’s not perfect, but it’s the best we can do right now.
Is sentiment analysis part of this path of understanding? “Sure,” Nolan said, “and we have a natural language processing suite that is part of our Decision Center and [our] customers, especially in customer service. You still need to train things and historically the process has been slow”.
Pega’s models are not designed to interpret a client’s emotional state as he tries to interpret. The problem with pure sentence patterns is that they are often very short-sighted; they don’t have the rest of the context for that person’s situation. And how do good models feel? It’s nice to know if someone is happy or sad, but it’s hard to act. Home.
Asked about Pega, Stepanova said: “ In the real-time interaction space, much of what they do is focused on managing client cases, because there is an individual conversation going on and we’re looking at the emotions wide open. ladder.
Send messages to the “intermediary”
For further analysis, we turn to Seth Grimes, chairman and chief advisor of Alta Plana, former organizer of a series of sentiment analysis symposia recently convened for the CX Emotion Conference in London.
“Companies have been conducting business analytics with reliable offerings for over a decade,” he said. “The most successful use cases I’ve seen were specifically about customer experience, not marketing. Customer experience is more like reflection, analysis, research, and so on.
Based on research conducted by Net Promoter Scores, Grimes recognizes evidence that it is effective and rewarding to communicate “between people” rather than trying to reverse negative attitudes. “If you can identify positive people, you can find sales and cross-selling opportunities.” This was in line with Iterable’s proposal, as described by Stepanova.
Additionally, Grimes said that including a machine learning component in sentiment analysis would differentiate an offering like Iterable from companies still using rule-based methods (in machine learning, the model corrects itself, possibly with some human supervision).
Empathy is the keyword
Whether it’s consumer sentiment or emotion, or the brand’s empathy with its customers, we’ll be talking a lot more in 2021. The question goes beyond the second-best marketing message of how brands generally find the right tone of voice.
“In the last ten months,” said Nolan, “the whole world has changed. Brands have gained analytical power over the years. Suddenly the world changes and all of our behavior changes, and they are just like these models.
The challenge is to connect with people whose worldviews may have changed this year. “Unless you have the resources to do it,” Nolan said, “it’s hard to know what to say and when to say it.”